Paulo Jan
Forum Replies Created
-
Paulo Jan
February 3, 2008 at 6:48 pm in reply to: Re-interlacing footage at the broadcasters request: normal or not?[Larry Asbell] “And stop saying “because both fields looked the same (as they had to, since the original footage was progressive).””
Sorry :-). What I meant is that, when looking at the prog. footage field by field in Avid or After Effects, there’s movement only every two fields; in other words, both fields in a frame are the same (or at least, Avid and After Effects display them as the same). Since that’s the method we usually use to examine field information, I tend to think in those terms.
Paulo.
-
Paulo Jan
February 3, 2008 at 6:37 pm in reply to: Re-interlacing footage at the broadcasters request: normal or not?Doh! Of course. In Walter’s case, he’s starting from a 720p60 (emphasis on “60”) source, which of course makes it much easier to interlace it later into a 1080i60 (i.e. 30fps) master. I was thinking of it from my point of view, where I have 25p footage that I have to convert somehow to 25i. Sorry, I hadn’t realized about this.
And, arc nevada: I’m not trying to fight any battle. As I said in my first post, we’ve already make arrangements to start shooting in interlaced and to try to “re-interlace” the progressive footage we already shot. I posted here just to gather some opinions, to see whether it’s the guy in the TV channel who is on crack or it’s me.
-
Paulo Jan
February 3, 2008 at 3:41 pm in reply to: Re-interlacing footage at the broadcasters request: normal or not?Oh, that’s what we did too. What we delivered was an interlaced DVCPRO, but the guy at the station was still complaining that it wasn’t really interlaced, because both fields looked the same (as they had to, since the original footage was progressive).
In your case, the AJA Kona can cross-convert from 720p to 1080i, but the interlaced result… is it “real” interlaced? In other words, if one were to look at it field by field, would each field look different from the next one?
Paulo.
-
Paulo Jan
February 2, 2008 at 10:11 pm in reply to: Re-interlacing footage at the broadcasters request: normal or not?Yep, we’re shooting 25fps, using the HVX-200’s “576/25p” mode. When watching the footage in Avid or After Effects field by field, I don’t see any resolution loss, just movement advancing every two fields.
-
Paulo Jan
February 2, 2008 at 9:42 pm in reply to: Re-interlacing footage at the broadcasters request: normal or not?First of all, thanks to both you guys for confirming that I’m not crazy. Now, I think this guy did know what “progressive” means, he just didn’t want it, for whatever reason. And yes, I thought about asking him: “so, when you guys show ‘Saving private Ryan’ in your channel, are you going to ask Spielberg to reshoot the battle sequence in DV just so that the source material comes interlaced?”. But I preferred to be diplomatic…
To be fair, we aren’t shooting progressive for the “film look”, but for a variety of technical reasons, and our show isn’t a high-end doc or anything, but an hour long magazine (don’t know if it’s how it’s called in English: sort of current-affairs, cultural events, etc. features), so arguably we should be shooting in interlaced, but I don’t really think it’s such a big deal: after all, our show isn’t going to be integrated anywhere else as Ben suggested (it’s a stand-alone), so it’s not as if there’s going to be a huge difference in looks (what I mean is: there isn’t going to be the kind of abrupt change Ben is talking about. I didn’t mean that there wasn’t difference at all between the progressive and the interlaced look).
Lastly: how do you reinterlace progressive footage? The solution I found was FieldsKit, which has that option. From the quick tests I did, it seemed to do a reasonable job, but of course, I’m afraid there *might* be some quality loss…
Paulo.
-
Paulo Jan
January 29, 2008 at 9:43 pm in reply to: Flare/stroke moving through a path: how to change speed?Thanks a lot for your answer! Unfortunately, the layer (and the flare”s movement) is 3D… and Write-On’s brush position seemd to be only 2D. Am I missing something, or is there some other way?
(I’ve just tried another way I just thought of: precompose the flare and the stroke, move the precomp into a new composition and use Time Remap to manipulate the speed. Unfortunately, it didn’t work in my case: the layer where the stroke takes place has a non-standard size (2000×1500), and precomposing it and replacing it with a 720×576 one messes up the camera movements that I had already set up).
Paulo.
-
You say that in Premiere, the PC never asks you for rendering; if it can’t keep up with a sequence or effect, the playback just becomes pixellated.
The equivalent of that in FCP is the “Unlimited RT” option. Set your timeline to that instead of to “Safe RT”, and you’ll see that the computer doesn’t ask you for renders; it will just skip frames when it can’t keep up.
This said, FCP’s scrolling text effect is a real dog; I’ve never understood why something as simple as scrolling text through the screen requires so much render time…
-
I might be wrong, but I think that the reason Apple advises this is so that the memory can be accessed in Dual Channel mode, which (in theory) doubles the memory bandwidth of the computer. Now, this said, there is some controversy in homebuilt PC forums (and these days Macs are just PCs with fancy cases 🙂 ) about how much performance gain Dual Channel actually buys you; I’ve read quotes that say that it isn’t much more than a 10% speed increase in practice.
Personally, I don’t have any first-hand experience about it, but I’ve always been curious about it, so when we upgrade our MacPros at work I’ll be sure to run some benchmarks.
Paulo.
-
But, doesn’t the manual say that “Export to Compressor” is the recommended option to get the most quality? The explanation I’ve always read was that, with a QT Reference, Compressor would be taking the render files, which would be already compressed into DV (or DVCPRO or whatever) and reencoding them again, while with “Export to Compressor”, Compressor would be receiving the raw, uncompressed render output from FCP, which would save a step.
The explanation makes sense to me, but I’ve always been curious about whether the difference in quality is actually worth it, or if it’s even visible. Would you say that the quality difference is negligible?
-
Well, yes, but the thing is that I have used clips with much longer filenames than the ones my colleague was using, and I’ve never had a problem. Then again, I captured those clips from tape, using Log & Capture. I’ve been searching around here and haven’t found anyone else complaining about this bug either. Can anyone confirm if Log & Transfer has indeed problems with long clip names?